Poking the Sloth

by Kieran Healy on July 17, 2007

I was going to pass over this, but I am a shallow person. Fresh from “schooling me”:https://crookedtimber.org/2007/07/15/dept-of-being-savaged-by-a-dead-sheep/ on the treatment of outliers, Megan McArdle has expanded her ambition and now “takes Cosma Shalizi to task”:http://www.janegalt.net/archives/009901.html for his “bizarrely beside the point” “views”:http://cscs.umich.edu/~crshalizi/weblog/495.html on the heritability of IQ, the statistical estimation and interpretation of _g_, and his failure to understand the analytical methods of “the serious IQ guys.” Megan may not be aware that I “taught”:http://www.stat.cmu.edu/~cshalizi/754/ “Cosma”:http://www.santafe.edu/profiles/?pid=236 “what little”:http://www.cscs.umich.edu/~crshalizi/prob-notes/ he “knows”:http://www.cscs.umich.edu/~crshalizi/research/ about statistics. He’s also much nicer than me. So she’ll have no trouble disposing of him.

_Update_: Yeah, on second thoughts I should have just passed over this.



Walt 07.17.07 at 8:03 pm

Given that Megan is an economics blogger, isn’t she out of line commenting on another area of social science? That’s my understanding of how these things work.


jw@jw.com 07.17.07 at 8:10 pm

No at all. Since she’s not really a qualified economist either, she’s the blogospheric equivalent of those horrendously annoying Holiday Inn Express “commercials”:http://www.ichotelsgroup.com/h/d/ex/1/en/c/2/content/dec/teaser/ex/1/en/lp/sstvcom.html.


Steven 07.17.07 at 8:22 pm

Prof. Healy-

While your snark is no doubt amusing (to some), it’s worth pointing out that McArdle most certainly did not say that Shalizi’s post represented a “failure to understand the analytical methods” of IQ researchers, but rather a failure to engage what she sees (correctly, in my estimation, although I am not positive on this) as the more common evidence used by (for lack of a better term) “IQ determinists,” that is, twin and adoption studies, and the like.

Exactly what is it in the post that would lead you to believe that McArdle is asserting that Shalizi doesn’t understand statistical methods?


mkl 07.17.07 at 8:34 pm

May I assert that the belief that IQ is heritable is strongly correlated, nay, concentrated in high IQ types without children, and notably absent among parents of several children, independent of the IQ distributions of those parents or their children? It’s nothing but a fond but selfish hope.


Nick 07.17.07 at 8:38 pm

I don’t know much about the heritability of IQ, but I’m baffled by Shalizi’s claim that accent is fixed after adolescence. I speak with an accent that is quite different than that of my parents. It has changed over time, and I can track those changes by people’s reactions when I talk. It was some English variant in junior high school, an English/American blend in college, and a nondescript midwestern accent with no obvious geographic markers now. I anticipate that it will continue to change if I stay here in the southeastern U.S.

I’ll bet that your accent is very strongly influenced by your unrelated peers in the same generation, significantly less influenced by your parents, and hardly influenced at all by your grandparents. Isn’t that pretty much the opposite of heritable?


P O'Neill 07.17.07 at 9:16 pm

No time like the present for a little Charles Murray. He was responding to letters about his big piece in Commentary magazine on IQ and Jewish people. One letter pointed out that measured Irish IQ has been rising rapidly.

David Quin’s recounting of the Irish story has many parallels with that of the Jews: the latent ability was always there, but for a long time conditions limited its expression. In earlier decades, Irish potential was hidden by lack of educational opportunities and the overwhelmingly rural character of Ireland. (Everywhere, IQ scores show a marked gradient from rural to urban areas, for reasons related partly to education and partly to the greater intellectual stimulation of urban areas.) But long before the recent Irish economic transformation, the reality of that potential had been proved by the Irish who had emigrated to the United States. As someone who has been visiting Ireland for forty years (my sister married a Limerick man), I should add that I have never doubted Irish verbal ability from the first time I sat in a working-man’s pub and overheard the conversational gymnastics going on around me.


Luis Alegria 07.17.07 at 9:45 pm

Mr. (or Ms.) Mkl,

I am a father of several children, and what I have observed leaves me little doubt that IQ is a heritable trait.


Walt 07.17.07 at 9:51 pm

I’m sorry to hear that, Luis.


ogged 07.17.07 at 9:55 pm

I am a father of several children


Mr. Alegria,

Maybe you and the other fathers can get together and nail down a number.


ogged 07.17.07 at 10:01 pm

Damn, Walt’s is funnier.


rilkefan 07.17.07 at 10:32 pm

“May I assert that the belief that IQ is heritable is strongly correlated, nay, concentrated in high IQ types without children”

That’s crazy. High IQ types with children very much believe that IQ is heritable – they are smart, their kids are smart, instant validation/family feeling. Mrs. R.’s side of the family falls squarely on the nuture side of things, but they still say Rilkekind is likely to be smart because of his genetic heritage and his environment.

‘A: Sure. In a more cautious mood, instead of saying “there is” I’d say “there could be, for all we know at present, which seems to be squat”. But, sure.”‘

This isn’t reflective of current scientific thought, is it (and isn’t the reluctance kind of bizarre)? It’s not reasonable to say the heritability of intelligence is 0% or 100%, for example, is it?


Watson Aname 07.17.07 at 10:39 pm

“May I assert that the belief that IQ is heritable is strongly correlated, nay, concentrated in high IQ types without children”

Perhaps high IQ type will account for the implication of regression to the mean, so would be unsurprised by having offspring who they consider to be somewhere between themselves and the mean?


Luis Alegria 07.17.07 at 10:45 pm

Mr. Ogged, Mr. Walt,

Good work !


Luis Alegria 07.17.07 at 10:48 pm

Mr. Rilkefan,

Correct, thats just what I mean. I have seen my kids and dozens of their peers grow up, and I know their parents. Smart, accomplished parents tend to have smart kids.


dearieme 07.17.07 at 10:53 pm

“Smart, accomplished parents tend to have smart kids”: quite so, but that could be caused by nature or nurture (or endless interactions of same). Hence the need for twin studies and whatnot.


BillCinSD 07.17.07 at 10:58 pm

What do you mean when you say smart, Luis? That’s the crux of the issue. I would contend that the twin studies may show more the correlation of doing well on various types of “IQ” testing than to intelligence.


notsneaky 07.17.07 at 11:01 pm

It’s not reasonable to say the heritability of intelligence is 0% or 100%, for example, is it?

Oh no, it’s the Laffer curve again!


paul 07.17.07 at 11:05 pm

Is this a demonstration of the First Law of Holes? Or just more proof of the Economist’s sad decline?


Megan McArdle 07.17.07 at 11:06 pm

a) I wasn’t critiquing statistical method, or saying anything at all about statistics, except that they don’t seem quite pertinent (which is the provice, AFAIK, mostly of psychologists, not sociologists, but I might well be wrong). I was just saying that I didn’t see how the evidence that accent is both heritable and plastic, refutes the evidence the IQ people say they have from adoption and twin studies that IQ is both heritable, and not very plastic short of extreme material deprivation. But perhaps I am missing something that studying statistics under Professor Healy might teach me?

b) I have no particular opinions about the roles of genes vs. environment, except that both sides tend to overstate their evidence quite a bit, and the race nuts repulse me.


rilkefan 07.17.07 at 11:18 pm

19 b) seems entirely correct to me.


John Quiggin 07.17.07 at 11:22 pm

As we discussed here at length a while back, both height and IQ have shown themselves surprisingly plastic in environments that could scarcely be described as involving “extreme material deprivation”.


Luis Alegria 07.17.07 at 11:25 pm

Mr. Billcinsd,

I am speaking of personal observation, not of academic studies, in response to Mr. Mkl’s assertion.


Martin James 07.17.07 at 11:27 pm

We should let the market settle this. We’ll take sperm from fathers of various IQ levels and eggs of women with various IQ levels and place them in homes of couples with various IQ levels.

Then people can but a future on the IQ at age 5, 10, 15 and 20 of the pool made up of the various combinations. For example, you have a biological father at 95, biological mother at 115 raised in a family of mother 100 and father of 105 and then we establish a market that pays say $10 times the IQ points of the child.

This will give us a measure of what people believe. Those that think IQ is less heritable can make big bucks shorting the high biological parental IQ pools and vice versa.

That’s one of the best discoveries of economics, the more money it costs people to hold false opinions(for example racism), the less likely they are to hold them.


rilkefan 07.17.07 at 11:28 pm

“Smart, accomplished parents tend to have smart kids.”

Well, smart, accomplished (assuming this is a reasonable proxy for ‘provides a good learning environment at home/encourages studying/passes down genes for studiousness’) parents tend to have quite smart kids; but s!a and !sa parents less so, and !s!a parents even less – isn’t it the informed conventional wisdom that nature and nurture are both important? Ridley’s book on the subject came out more than a few years ago and seemed quite sensible to me – I think he said the data was most consistent with 50% heritability (vs a much higher rate for personality).


Megan McArdle 07.17.07 at 11:30 pm

I’m really pretty disinterested in debating the actual nuts and bolts of IQ heritibility, because as you say, the level of vitriol and bullshit from both sides just keeps increasing. My main point was that to the extent that IQ is interesting to people who are not simply looking to vindicate their preconceptions about race and gender–and I actually think the most interesting questions lie well outside those two minefields–that post doesn’t really seem to shine any light on the question. It just seems like refuting a straw man, the raving loon who thinks that IQ is 100% genetic. There are such people, but since they don’t seem to be pervious to reason there’s not really much point engaging them. But I wasn’t taking on that post because I wanted to argue that IQ isn’t plastic; I don’t know the answer to that, and haven’t examined the question closely enough to have an opinion.


Luis Alegria 07.17.07 at 11:31 pm

Mr. James,

Why the betting ? That scenario sounds like it would work very well as a plain old experiment – not that it is likely to be attempted outside of North Korea, maybe.


Luis Alegria 07.17.07 at 11:35 pm

Mr. Rilkefan,

I can only recommend the Gene Expression blog, where they are up on all the latest literature and know all about genetics, which is way beyond me. From all that I can tell there seems to be a consensus somewhere upwards of 50%.


Martin James 07.17.07 at 11:41 pm


Betting is good for several reasons. It tells us what people believe is true not just what is true plus we don’t need to quibble about all the confounding factors in the experiment the market will price that in.

Plus everybody’s happy – they all think they are going to win!

About North Korea. I saw a television show about the place a few years ago. They showed a lot of very beautiful families. Either the standards of beauty used to put people on TV are universal or all that communism and isolation are very good for the health.


rilkefan 07.18.07 at 12:00 am

So my question remains, is it reasonable for Shalizi to write about the non-null hypothesis, “for all we know at present, which seems to be squat” instead of, “the consensus view in the field is that h is 50%, but I’m skeptical about whether the required correlation matrix is right because blah” or whatever.


engels 07.18.07 at 12:09 am

Ogged: Walt pwned you.


Kieran Healy 07.18.07 at 12:09 am

I wasn’t critiquing statistical method, or saying anything at all about statistics, except that they don’t seem quite pertinent

Megan, the plausibility of twin studies, and the IQ enterprise as a whole, is tied up with the nature of _g_, its measurability/quantifiability, and its connection to our folk concept of intelligence. And this is very much a question of statistical theory, and it’s in that context that Cosma’s post should be seen — he is challenging the idea that _g_ is the sort of entity that some think it to be, and hence questioning its usefulness as a measure. You proposed twin studies vs general observational data as th answer, but this presupposes that prior questions about the measurability of IQ are resolved — it’s not a solution to those problems.


Barry 07.18.07 at 1:24 am

“I can only recommend the Gene Expression blog, where they are up on all the latest literature and know all about genetics, which is way beyond me. From all that I can tell there seems to be a consensus somewhere upwards of 50%.”
Posted by Luis Alegria

IIRC, the founder is an admirer of ‘The Bell Curve’, which invalidates anything that he might say.


agum 07.18.07 at 1:45 am


Steve Sailer to thread.


walt 07.18.07 at 1:51 am

In this whole thread, only Engels has the courage to tell the truth.


Jon H 07.18.07 at 2:44 am

Smart, accomplished people are confident that their spawn will be smart and accomplished, until the missus squeezes out a Shrub dud or two.

At which point the parents become devoted believers in everyone being special and talented in their own way, even if they aren’t ‘smart’.


Jon H 07.18.07 at 2:49 am

” I have seen my kids and dozens of their peers grow up, and I know their parents. Smart, accomplished parents tend to have smart kids.”

A tendency is not a guarantee.

Given that couples don’t poot out 100 kids, it only takes one dud child to really sink your brood’s average IQ. Given that so many couples have only one or two children, it’s entirely possible that a smart couple will have only low-IQ kids.


s.e. 07.18.07 at 3:54 am

It’s like listening to the teachers’ pet at a third rate Catholic school arguing with the computer geek with Aspergers syndrome.
Jesus christ. Grow up the fuck up. If Shalizi is your idea of intelligence we’re all doomed.

“Jane Galt” doesn’t even fucking exist. Argue with someone who isn’t an absolute idiot, or ignore them.


Kieran Healy 07.18.07 at 4:17 am

Jesus christ. Grow up the fuck up

Why don’t you try calming the fuck down.


Luis Alegria 07.18.07 at 5:21 am

Mr. Jon H,

Of course. Now how many parents will tend to have the experience you describe, vs the others ? There is such a thing as a statistical truth.


Luis Alegria 07.18.07 at 5:28 am

Mr. Barry,

I have no settled opinion on “The Bell Curve”, but I will just note that your grounds for dismissal (of GE and its owner through association with Murray) are thoroughly unsound both scientifically and logically (ad hominem, etc.). In the area of political rhetoric of course there are no such rules.

Which neatly illustrates one of Razibs, and GE’s, typical points about the absurd politicization of the subject.


John Quiggin 07.18.07 at 6:35 am

Miss Luis,

As a few people have pointed out already, “Mr Firstname” is not an appropriate form of address in most English-speaking societies, particularly when you have no idea whether the person concerned goes by some other title, or no title at all.


SG 07.18.07 at 6:49 am

Mr. Luis

I`m sure you go to the trouble to check every fact and conclusion avowed communists present you with. I`m sure you never dismiss them out of hand.

This is why, for example, when discussing global warming, you never dismiss it as “greenery” intended to further the cause of communism. Such a method of disputing a theory would be unscientific and ad hominem, wouldn`t it?


Zora 07.18.07 at 7:35 am

IMHO, resistance to anything that threatens the dogma of “no essential genetic differences” is a defense of meritocracy.

For those who believe that Western society IS meritocratic, anything that undermines the notion that the outcomes are FAIR is threatening. If we don’t all have an equal chance at winning the intelligence lottery, then a meritocracy becomes a caste-based society.

If you don’t believe that society is meritocratic, but that it should be, then you must postulate that changing social conditions could make the race FAIR. If some human populations (say, Ashkenazi Jews) are naturally more intelligent than others, then leveling the playing field is still going to produce a caste-based society. That is, if you believe that power and wealth should be handed out to the “winners”. But who but the philosopher-kings are going to warm to that idea?

There’s no compelling reason to believe that evolutionary pressures stopped operating on humans at some point in the distant past. It’s quite possible that there are differences in intelligence or intelligences between some populations. However, we can’t know for sure, since we don’t have a level playing field.

I suggest a grand experiment: let’s level the playing field. Make sure every human being has adequate nutrition and health care, top-notch schooling, and social encouragement to learn and excel. Let’s abolish hierarchy, so that no one comes from a denigrated class. Once we’re there, then we can check for variations in native intelligence :)

(A variation on Habermas.)


abb1 07.18.07 at 8:39 am

Smart, accomplished parents

Oxymoron. Had the smart ever managed to accomplish anything, the world wouldn’t be a sorry place it is now.


shaftesbury 07.18.07 at 9:41 am

comment threads on IQs: where all the children and their parents are above average.


soru 07.18.07 at 10:35 am

‘it’s entirely possible that a smart couple will have only low-IQ kids.’

It’s raining and quite chilly today. Someone ring up Nature and get them to hold the front page slot for a decisive invalidation of all that global warming stuff.


tom 07.18.07 at 10:53 am

“Why don’t you try calming the fuck down.”

Heh, says the original Mr ‘Knickers in a Twist’.


Barry 07.18.07 at 10:57 am

John, that’s one of Allergica’s traits – and not his worst.

Allergica, when somebody with alleged expertise in a field (Razib, probably, although they like pseudonyms) endorses flat-out junk science in their field, then they’ve demonstrated either incompetance or dishonesty. It’s perfectly logical and quite decent method; I can, however, understand why you might not like it.


Kieran Healy 07.18.07 at 12:29 pm

Heh, says the original Mr ‘Knickers in a Twist’.

Let me see if I can remember what comes next — “I know you are buy what am I” maybe? I should have listened to my original feeling and ignored the whole thing.

Comments on this entry are closed.